A small experiment with players around Gold 3 to Platinum 3 found that teams with the higher-ranked tank often won. How much can we take from that, and what does the tank swap result tell us?
How much does the tank matchup affect wins in Overwatch?
Answers
The higher-ranked tank's team won 71% of the matches where the tanks had different ranks. That is an interesting result, but it doesn't establish how much the tank alone caused those wins.
The teams were arranged to have equal average ranks for the first match, which helps, but an average can hide differences between individual players and roles. There were also only 40 participants, some knew each other, and the rank range was fairly narrow. The result is a clue worth testing, not a win rate to expect in ordinary ranked games.
Reply to this answer
The swap result is less clear than it might look. After the tanks switched sides, the first game's winning tank won the second game 52.5% of the time. That doesn't mean the tank has only a small effect, because the second teams were no longer balanced in the same way as the first.
Also, winning the first game isn't the same thing as being the higher-ranked tank. The two percentages answer different questions, so I wouldn't compare them as if they measured the same effect.
Reply to this answer
I'd want a bigger test with more players and repeated tank swaps across different team combinations. Keeping track of who has played together and accounting for other role ranks and playtime would help too. The discarded matches are worth documenting, since the source doesn't say how many were removed.
Reply to this answer
Add your answer
Answers are reviewed before publication.
DpadDiagonal_6W:
The players also took part in multiple matches, so the results aren't necessarily 40 completely independent observations.